evaluation of james lip pressure method for low flow rate ... · proceedings world geothermal...
TRANSCRIPT
Proceedings World Geothermal Congress 2015
Melbourne, Australia, 19-25 April 2015
1
Evaluation of James Lip Pressure Method for Low Flow Rate Geothermal Well:
ML-5 Case Study
Robi Irsamukhti, Alfianto Perdana Putra, Novianto
Supreme Energy
Sudirman Central Business District (SCBD) Lot. 9
Jl. Jend. Sudirman Kav. 52-53, Jakarta 12190 – Indonesia
Keywords: James lip pressure method, production test, low flow rate geothermal well, muara laboh.
ABSTRACT
Currently, there are six exploration wells in Muara Laboh field. One of these wells where James lip pressure method has been
applied is ML-5. The interesting feature of this well is the low flow rate characteristic that has been demonstrated after drilling.
This feature is suspected to have influenced the accuracy of James lip pressure method.
Beside James lip pressure method, other intensive well testing surveys had been conducted at ML-5 during flow performance test,
including tracer flow test (TFT). The TFT is a reliable method able to give accurate analysis for flowing enthalpy and mass flow
rate of geothermal wells.
The work is accomplished to evaluate the accuracy of James lip pressure method for ML-5. Results of the method are analyzed and
compared to the results of TFT method. Possibilities of error are identified. Several aspects that may contribute to the errors are
elaborated. The correction process is conducted and sensitivity analysis is performed.
Case of ML-5 shows that relative error of James lip pressure method falls in a 10 – 14 % error margin relative to the result of TFT
method. The work clearly demonstrates that accuracy of the method is very sensitive to the measurements on brine water level at
the weir box. Further investigation shows that it is still possible to maintain the accuracy in the range of ± 2 % error margin by
improving the measurement on brine water level at the weir box. Furthermore, the work also suggests that TFT survey is important
as benchmarking to guarantee the reliability and accuracy of James lip pressure measurement.
1. INTRODUCTION
Muara Laboh is located in South Solok Regency, West Sumatera, Indonesia. This place is a geothermal working area which is
being explored by Supreme Energy. Currently, 6 exploration wells have been drilled in Muara Laboh to prove and delineate the
existence of geothermal system beneath the Muara Laboh area. Some of the wells have been discharged and flow performance test
has been performed.
James lip pressure method has been applied in Muara Laboh field to assess performance of the wells during discharge test. One of
these wells is ML-5. There is interesting feature of this well which needs to be investigated. This feature is the low flow rate
characteristic that is suspected to have influenced the accuracy of James lip pressure method.
James lip pressure method is a well-known technique to estimate the flowing enthalpy and total mass flow rate from a discharging
geothermal well. Enthalpy and total mass flow rate are required to calculate dryness of the fluid, steam mass flow rate, and estimate
productivity of the wells at a given well head pressure. This method is derived from the empirical experiment by James (1966).
James lip pressure method has been widely used in the geothermal industry. This method is preferred since it is cost efficient and
easy to be set in the field. The simplicity in both hardware and instrumentation are the advantages of this method. Furthermore, the
method is able to provide real-time monitoring on the well production data. James lip pressure method is a suitable method for
long-term production test of new exploratory wells when compact surface facilities are not available yet. These features are the
primary consideration to apply James lip pressure method in Supreme Energy fields.
Accuracy of the James lip pressure method has been frequently reviewed. Results are usually within 5% error margin for both total
mass flow and enthalpy. However, another study by Karamaraker and Cheng (1980) showed that this method gives the results with
8% error margin (Grant, 2011).
The objective of this paper is to evaluate the accuracy of the lip pressure method for low flow rate geothermal wells with case study
from ML-5. Indeed, when fluid enthalpy and steam mass flow rate are relatively low, small difference between calculated and
actual value will result in significant impacts on estimating the well productivity.
2. REVIEW OF LIP PRESSURE METHOD
James lip pressure method is implemented to geothermal wells being discharged to atmospheric conditions. The atmospheric flash
tank (AFT) is used to separate brine and steam fraction. Steam will be flashed to the atmosphere and brine is flowed to the weir
box. Short-closed pipeline is built to make the connection between well head and AFT. The pressure is measured at well head point
and at the extreme end of the pipeline as it enters the AFT which is called “lip pressure”. Brine flow rate is measured at the weir
box.
Irsamukhti et al.
2
Grant (2011) presented the procedures to calculate flowing enthalpy and total mass flow rate by using James lip pressure method.
The measured parameters are well head pressure in bar absolute, lip pressure in bar absolute, and height of brine from the base of
triangular-notch in the weir box which is usually measured in centimeter in the fields. The design parameter is the cross-sectional
area of the “lip-pipe” in cm2.
Relation between total mas flow rate (kg/s), enthalpy (kJ/kg), cross-sectional area of the lip pipe (cm2), and lip pipe pressure (bar
absolute) is described by James formula as below:
where W is the total mass flow rate, H is the fluid enthalpy, A is the cross-sectional area of lip pipe, and P is the lip pressure.
Relation between total mass flow rate and brine flow rate which is measured at the weir box is written as:
where the parameters with apostrophe are evaluated at the atmospheric pressure.
Substituting the equation (2) into equation (1) gives:
Grant (2011) denoted the left side of equation above as Y and made the relation between Y and enthalpy for atmospheric pressure of
1 bar as below:
Once the flowing enthalpy has been known, the total mass flow rate can be calculated using Eq. (2).
Assumed atmospheric pressure of 1 bar, procedure for calculating the enthalpy and total mass flow rate by using James lip pressure
method are:
1. Calculate Y using left side of Eq. (3)
2. Calculate enthalpy of fluid using Eq. (4)
3. Calculate total mass flow rate using Eq. (2)
In order to calculate Y value, the brine flow rate at the weir box (W’) must be calculated first.
3. CALCULATING BRINE MASS FLOW RATE BY USING THE TRIANGULAR-NOTCH THIN-PLATE WEIR
Shen (1981) proposed the semi-empirical correlation to estimate flow rate from the triangular-notch thin-plate weirs. This
correlation can be applied to estimate the brine flow rate at the weir box condition in order to calculate Y value in Eq. (4).
The proposed equation from Shen (1981) is written as below:
√
⁄
where is the flow rate in m3/s, Ce is the discharge coefficient, g is the gravity acceleration in m2/s, is the notch angle, he is the
effective head in m and defined as:
where h is the level or height of brine from the base of triangular-notch in the weir box and is the head-adjustment factor to
accommodate effects of viscosity and surface tension. Equation (5) is described as the Kindsvater-Shen equation in the international
standard (International Standards Organization, 1975) as quoted in Shen (1981).
Shen plotted the average values of Ce versus notch angles as is illustrated in Fig. 1. Based on the plotting, we have made the
approximated correlation between Ce and which can be written as:
Irsamukhti et al.
3
Fig. 1. Values of as a Function of Angles
Fig. 2. Values of as a Function of Angles
Shen plotted the values of versus notch angles as is illustrated in Fig. 2. Based on the plotting, we have made the
approximated correlation for as function of which can be written as:
The Eq. (7) is valid for degree with error margin less than 1%. Then Eq. (8) is valid for degree with
error margin less than 8%. Having calculated value using Eq. (5), brine mass flow rate at the weir box can be calculated as
below:
where is the density of brine water which is evaluated at atmospheric pressure.
When brine mass flow rate at the weir box has been known, then Y value in the James lip pressure method can be calculated as:
4. APPLICATION OF LIP PRESSURE METHOD IN ML-5
ML-5 is one of the exploratory wells situated in Muara Laboh geothermal prospect area. This well was spudded on 22nd of June
2013 and completed on 29nd of July 2013. A completion test and intensive well testing surveys have been done. Subsurface
information about ML-5 had been well characterized.
On 22nd of September 2013, ML-5 was switched to the flowing condition for the first time. Air injection technique was performed
to initiate the flow. This well has been discharged for 88 days or almost 3 months before being switched to shut-in condition on
19nd of December 2013.
Horizontal discharge method has been implemented in ML-5 to reduce effects of direct exposure from geothermal fluid to
environment. Using this method, two-phase fluid has been flowed to the atmospheric flash tank (AFT) through short-closed
pipeline. In the AFT, steam has been flashed to the atmosphere and brine has been flowed to the weir box.
Three are three parameters that have been measured continuously during flow performance test of ML-5. These parameters are well
head pressure, lip pressure, and height or level of brine water in the weir box. Well head pressure and lip pressure has been
measured by using digital pressure gauges. Level of brine in the weir box has been measured manually by using measurement set
which had good calibration in the field to make sure that the reading is correct.
Irsamukhti et al.
4
Fig. 3. Discharging of ML-5 through AFT on October 9, 2013
Fig. 4. Brine in Weir Box on October 9, 2013
Grant (2011) has made the relation between enthalpy H and Y values for atmospheric pressure of 1 bar as written in Eq. (4).
However, related to high elevation, the atmospheric pressure in ML-5 is less than 1 bar.
Based on the elevation of ML-5, the calculated atmospheric pressure is 0.84 bar. Then we have calculated the modified correlation
between enthalpy H and Y values for atmospheric pressure of 0.84 bar as is presented in Fig. 5.
Fig. 5. Relation between H and Y for atmospheric of 0.84 bar and kJ/kg
The modified correlation between enthalpy H and Y values for atmospheric pressure of 0.84 bar can be written as:
where,
for
for
Irsamukhti et al.
5
for
for
The correlation which is described in the Eq. (11) is valid under the enthalpy conditions quoted above with error within kJ/kg.
This correlation was used in the James lip pressure method to calculate the enthalpy and mass flow rate of ML-5.
5. EVALUATION AND METHODOLOGY
Tracer flow test (TFT) method has been conducted in ML-5 in order to obtain the accurate estimation for enthalpy and mass flow
rate. TFT method is capable to determine the enthalpy and mass flow rate of the two-phase flow stream of geothermal well in the
production line. The method requires the injection of separated chemical tracers for the liquid and vapor phase into flow stream.
Samples are taken downstream from the tracer injection point. Mass flow rate of each phase can then be calculated, and by knowing
the pressure in the pipeline, enthalpy can also be calculated.
TFT method is considered able to give the most reliable and accurate estimation for the enthalpy and mass flow rate. Hirtz and
Lovekin (1995) were performed field validations of TFT results based on experiences in Roosevelt, Salton Sea, and Heber
geothermal field. Validation process was done by using orifice meter measurement as reference. The validation result has shown
that greatest deviation is 4.37 % for steam flow rate, 4.61 % for liquid flow rate, and 4.10 % for enthalpy.
In Muara Laboh geothermal field, TFT had been implemented as benchmarking method to check the accuracy of measurement
from James Lip pressure method. In high productivity well that had been drilled in Muara Laboh, the difference between TFT and
James lip pressure method is in the range of ± 5 %. However, different result was found in low productivity well.
In case of ML-5, two TFT surveys have been performed to get the accurate values of the enthalpy and mass flow rate and at the
same time to validate also the results of James lip pressure method. These TFT surveys were performed on October and December
2013. Result of these TFT was also supported by result from pressure – temperature (PT) logging. It had been found that TFT result
has shown good consistency with PT logging result. Enthalpy from the TFT result was close to the feedzone temperature from PT
logging data.
Enthalpy and mass flow rate which have been calculated by using James lip pressure method are compared to the TFT result as
reference. Deviation of James lip pressure method to the TFT results is calculated as relative error which is defined as:
| |
where is the relative error, denotes the quantity from the lip pressure method, denotes the quantity from the TFT.
6. RESULTS
Two TFT were performed on 20nd of October 2013 and on 12nd of December 2013. Calculated enthalpy and total mass flow rate
from James lip pressure method and results from the TFT are presented together in Fig. 6 and Fig. 7 respectively.
Fig. 6. James Lip Pressure Results - Data October 2013
Irsamukhti et al.
6
Fig. 7. James Lip Pressure Results – Data December 2013
As can be seen in Fig. 6 and Fig. 7 above, there are deviation of James lip pressure results to the TFT results. Calculated error for
enthalpy and total mass flow rate are and on data 20nd of October 2013 and and on data 12nd
of December 2013 respectively. These results show that average error for enthalpy falls in value of and average error for total
mass flow rate falls in value of . Considering the average deviation of TFT is 4 % for enthalpy and 4.5 % for mass flow rate as
presented in Hirtz (1995), it can be calculated that in case of ML-5, the average error for enthalpy is and average error for
total mass flow rate is . Furthermore, it also indicates that accuracy which has been obtained in this case is less reliable than
accuracy that has been mentioned in Grant (2011) where the James lip pressure method usually gives the error margin within value
of .
The errors that are stated above are the relative errors which are calculated relative to the TFT result. One possibility could be that
TFT result not reliable. Regarding this issue, TFT result then validated with pressure-temperature (PT) survey under shut-in
condition and wellbore modeling of pressure-temperature-spinner (PTS) survey under flowing condition. Previous PT shut-in
survey has been done on 20nd of September 2013 and PTS shut-in surveys have been conducted on 30nd of October 2013 and 17nd
of December 2013. Calculated feedzone enthalpy based on these surveys is falls in the temperature range of 220 – 220 0C under
condition of compress liquid. Enthalpy from the TFT is also falls in this range. It confirms that TFT result is reliable.
Based on the analysis, several aspects might have contributed to the error of James lip pressure method result in the case of ML-5.
The aspects have been identified as below:
1. The significant amount of non-condensable gas (NCG) in the wellhead may have led the calculated enthalpy to a less
reliable value.
2. The unstable well condition may have led the calculated mass flow rate to a less reliable value.
3. Unreliable pressure gauges measurements.
4. Deposition of mineral inside the U-pipe of measurement set in the weir box.
5. The manual reading process on the weir box may have led the reading of brine water level to a less reliable value.
Geochemistry survey had been conducted on ML-5. Non-condensable gas (NCG) was determined from the survey by analysis of
samples of steam condensate from the two-phase flow line. The analysis yield NCG concentration of 0.025 – 0.037 weight %.
These concentrations were corrected to total discharge. This correction was performed because gas concentration in steam will vary
with sampling pressure but in total discharge will not. Furthermore, calculated NCG flow rate at the wellhead of ML-5 is showed in
Fig. 6 and Fig. 7. It clearly shows that NCG flow rate is very low (less than 0.05 kg/s) and tends to be zero under flowing condition.
Therefore, the NCG effect can be excluded from the list of possible aspects that may contribute to the error.
The unstable well condition can lead the calculated mass flow rate to a less reliable value. However, in the case of ML-5, even the
well has shown a slight increase of performance every month, it has been observed that reading of wellhead pressure on 20nd of
October and 12nd of December 2013 stabilized at certain value. Therefore, contribution of this aspect to the error can be ignored.
The other possibility that may have contributed to the error is the unreliable pressure gauge instruments. However, based on the
field investigation, it was found that the digital pressure gauges which were used at the well head and lip-pipe of ML-5 had given
reliable measurements. There was no significant contribution from these pressure gauges to the errors.
Mineral deposition if occurred inside the U-pipe of measurement set in the weir box can obstruct and reduce the inside diameter of
the pipe. Based on the capillary principle, this occurrence will influence the height or brine water level in the measurement set. This
Irsamukhti et al.
7
phenomenon can lead to a unreliable brine water level measurement. However, in the case of ML-5, regular clearing had been done
to make sure there was no mineral deposition left in the U-pipe.
Further investigation has been addressed to the measurement of level or height of brine water from the base of triangular-notch in
the weir box. Brine water level in the weir box had been measured manually. Calibration for this set had been done and frequently
checked in the field.
Even during the calibration of measurement in the weir box, however, it had been observed that brine water level in the weir box
was never smooth. Brine water level in the measurement set had fluctuated and slight isolation phenomenon had been observed in
the reading. In the case of ML-5, reading had been simplified by take the average values of brine water level in the measurement
set. Based on this situation, it was considered that level of brine water measurement had the most potency to lead the result of
James lip pressure to less reliable values.
Fig. 8. Simple measurement set with U-pipe principle to measure the brine water level in the weir box
Sensitivity analysis is performed to investigate the impact of brine water level to the calculated enthalpy and total mass flow rate.
The process is done by adding relatively small values as corrections to the measured brine water level, as is presented in Fig. 9 and
Fig. 10, and then re-calculates the enthalpy and total mass flow rate by using those new values. Then results are compared to the
TFT results and relative errors are calculated by using Eq. (12). The results of sensitivity analysis are showed in Fig. 9 and Fig. 10
below.
Fig. 9. Sensitivity Analysis - Data October 2013
Irsamukhti et al.
8
Fig. 10. Sensitivity Analysis - Data December 2013
As can be seen in Fig. 9 and Fig. 10, adding small values to the measured brine water level in the weir box, until a specific number,
lead to the significant decreasing of the relative errors. Then, if this process is continued, the relative errors start to increase
consistently. Below of a specific number, the relative errors for total mass flow rate are less than enthalpy. On the contrary, above
of a specific number, the relative errors for total mass flow rate are more than enthalpy.
As can be seen in Fig. 9 and Fig. 10, there is one specific number where the relative error for enthalpy is exactly same with the
relative error for total mass flow rate. This number is considered as minimum correction value that can be reached to achieve the
optimum relative errors for both enthalpy and total mass flow rate at the same time. For data on 20nd of October 2013 illustrated in
Fig. 9, this minimum correction value of cm can give the optimum relative error of . Then for data on 12nd of
December 2013 is presented in Fig. 10, the minimum correction value of cm can give the optimum relative error of %.
Fig. 11. Corrected Results - Data October 2013
Irsamukhti et al.
9
Fig. 12. Corrected Results - Data December 2013
The minimum correction values that have been obtained above are implemented to re-calculate the enthalpy and total mass flow
rate of ML-5. These calculated enthalpy and total mass flow rate are presented in Fig. 11 and Fig. 12.
Fig. 11 and Fig. 12 clearly show that correction process that has been done to the measured brine water level has significantly
increased the accuracy of James lip pressure method in ML-5. This result indicate that if measurement of brine water level can be
precisely and accurately done in the field, more reliable calculated enthalpy and total mass flow rate can then be obtained by using
James lip pressure method with relative error within for both of them.
7. CONCLUSION
The conclusions that can be drawn from the accomplished work can be written as below:
1. James lip pressure method with standard measurements on brine water level in the weir box by using simple measurement
set, may lead the results to relative error margin of more than 10 %. In case of ML-5, we have obtained relative error margin
of 15 % for enthalpy and 11 % for total mass flow rate.
2. Accuracy of James lip pressure method is very sensitive to the measurements on brine water level in the weir box. Case of
ML-5 shows that correction that has been made was able to improve the accuracy of James lip pressure method until relative
error margin less than 2 %. This correction was done by adding correction factor of 1.3 – 1.6 cm to the reading of brine
water level in the weir box.
3. Accuracy of James lip pressure method is still possible to be maintained in the range of relative error margin around
by improving the measurement of the brine water level in the weir box. This is possible to be conducted by applying the
precision real time electronic measurement.
4. Benchmarking with other standard methods is important to be performed to guarantee the reliability and accuracy of the
results obtained. In case of ML-5, TFT method was used to provide a comparison to James lip pressure method.
ACKNOWLEDGEMENTS
The authors would like to sincerely acknowledge the following people: Julfi Hadi, Jantiur Situmorang, Muhammad Tamrin
Humaedi, Rudy Martikno, Alexandre Bacquet, and Alfiady, who have supported us in the writing of this paper. Furthermore, the
authors wish to acknowledge Supreme Energy which has permitted to publish this work. We are grateful to Subsurface Team in
Supreme Energy.
REFERENCES
Grant, Malcolm A., et al. 2011. Geothermal reservoir engineering – 2nd Edition. Academic Press.
Hirtz, P., and J. Lovekin,. 1995. Tracer dilution measurements for two-phase geothermal production: Comparative testing and
operating experience. Geothermal Resource Council Transactions, Vol. 19, pp. 563-568.
International Standards Organization. 1975. Liquid flow measurement in open channels using thin-plate weirs and venture flumes
(ISO 1438-1975).
Irsamukhti et al.
10
James, C.R., 1966. Measurement of steam-water mixtures discharging at the speed of sound to the athmosphere. New Zealand
Engineering 21 (10)
Karamaraker, M. and Cheng, P., 1980. A theoretical assessment of James’ method for the determination of geothermal bore
characteristic. Rep LBL-11498 (GREMP-12), Lawrence Berkeley Laboratory, California
Martikno, Rudy., et al. 2013. Evaluation of weirs calculation to estimate well capacity: a numerical study. Proceedings of 13th
IIGCE.
Shen, John. 1981. Discharge characteristics of triangular-notch thin-plate weirs. Geological Survey Water-Supply Paper 1617-B,
US Department of Interior.
Supreme Energy. 2013. Muara Laboh well testing reports and documents. Unpublished.